MegaSquirt 2 Tune

I’m in a difficult situation where I live in an area where only a very small pool of people will even consider helping tune my car. Mainly because I run an megasquirt 2 ECU. The fact is that no tuners are wiling to spend the time on it, most actually said bring it in and we’ll swap out your ECU. Now I’m of the mind set ‘if it ain’t broke don’t fix it’…

I have built the Ford Duratec 2.0ltr engine with a high lift cams, worked head, pocketed pistons and lightened crank. It has Jenvey 45s with staged injection running 200cc Bosch injectors on the head and 210cc Bosch injectors on the throttle bodies.

Now after the initial build I found one guy who was initially willing to help. He swore all though the tune at how he hated the ECU brand. But effectively what he did for me was 10-15 runs and sent me on my way with full throttle pulls on his dyno. At this time i was fairly happy as met the figures i was expecting. Although the torque curve was a bit wallowy.

The issue I’ve found is that on the road in most driving its struggles low down, considering the car is very light (560kg) it should I think do better.

I’ve attached a print of the dyno pull at full throttle, also a screen shot of the staged injection. Now its my understanding that the MS2 does not disengage the primary injectors when engaging the secondaries even at 100% so also attached a table of roughly my injectors capability at 100% efficiency (I understand these should be typically oversized). Now to add to my confusion the ignition table has a mid point drop, that makes little sense to me – I’ve considered its been purposefully put there for a bit of retardation if you lightly hoover on the throttle through a fast bend… shot of the ignition table attached. The more i look the wronger it feels!

To compound issues my fuel table reads upto 195 VE but my engine has no forced induction, I’m assuming this is to stretch the injectors to their max..

The aim of this is trying to make it more driveable, I get no real pull until both injectors reach 100% on the staged injection detailed below in CC;

This shows that you can see unless I’m at 95%tps and 5000rpm I’m not getting full injection capabilities and I believe that’s affecting my power and possibly underfueling me resulting in the engine running lean. Ive purchased some 440cc injectors (waiting on postage) to put on rail 1 and I will keep 210ccs on rail 2 to top up at high revs, but a little nervous about the change needed to the fuel table!

I have no dyno to check available to check, I have taken logs, reviewed them and ran VE analyse from the mega log viewer. Which made fuel cell changes prior to the shown below, which is interesting as its mainly where the cam comes in that its adding alot of fuel but reducing fuel prior. (On the table blue is richer red is leaner)

I've generated a new ignition table which feels a tiny bit more consistent and safer that the one im running on and the new VE map below with my injectors rails both running 100% other than when i lift from the throttle

Here is the new VE Table..

I've prattled on a bit above, pretty much sums up my learning so far, I need to find an affordable knocksensor. But I guess i'm looking for assurance that i'm on the right track. I still don't fully understand why my fuel table shows 100+VE on an NA engine. I dont want to plug a map in and go bang! :D It could be thats as good as the engine will give and the torque curve wont improve, but all online sources pretty much confirm my injectors are too small even when combined at 100% efficiency.

Thanks for any help or advice given - no matter how critical!

Do you have a wideband Lambda/O2 sensor fitted? Can you provide a log of a full and partial power RPM pull from 2000 to Redline in say 2nd or 3rd gear?

The forum won't let me attach an mlg file so i zipped them. There's two but they've been going through the mill on ve analyse.

Here's a snap shot, but i've made a change and switched both rails of injectors on through the staging. So its running through the autotune process and a bit of iteration. But to be honest ive been struggling with my RS232 cable dropping out which is a complete pain so keep losing complete logs. It is a bit rich still since switching on both rails but hopefully auto tune will bring that down.

I don't see what you are trying to do -- are you just driving around, or are you trying to hit cell targets and work on particular parts of the table? If so, you need to work on making your throttle position steady (control the speed with left foot braking) and give the wideband sensor time to respond (and the autotune time to make a difference). Go for a particular RPM target and try to hold the engine steady at that speed. Use more throttle and more brake to reach more of the table. If you have some long-hills you can climb that helps, too.

Personally, I would not enable the staged injection until the throttle position was greater than 50%, and then I would bring it on smoothly (say 0% at 20, and 40% at 50, 75% at 70 and 100% at 90. You will get much more response and OEM-like behavior using the injectors mounted in the head. Put Duty_Cycle and Duty-Cycle2 on the same graph, to understand when you need the secondary injectors, and how much.

Smooth and interpolate that ignition table, once above 5000 RPM and below 15% load, all those cells can be the same large advance (34-40 typically), the retarded values you have now might result in more exhaust noises unless you are cutting fuel in those cases.

If your RS-232 dropping is a mechanical issue, get jack screws for the connectors. If your computer is going to sleep and dropping the connection, then work on the energy settings to prevent that.

BTW - 270 hp from a 2L Duratec is in the range of what I expect, and that would certainly mean a VE well above 100.

Firstly these logs are just ones I've extracted where I have been just driving around.

Maybe foolishly I did think of I got 2 hours or so driving in I'd run ve analyse from the log viewer rather than auto tune whilst I'm driving and id have probably got alot of cells throughout the drive. I considered this probably as more reliable than auto tune given it makes the changes at the end rather than during...

As you've suggested would be better and I'll try and do that once I've sorted my connection issues, although it's tough on our busy roads.

I've a rs232 Bluetooth module coming. Which will hourly allow some consistent logging

With regard to the injection I'm waiting for larger injectors to turn up as I was under the impression 200cc weren't up to the job, my staged injection was how you said but I was experimenting to see how the car performed with all the injectors working effectively giving 410cc all the time rather than 400cc at 5k rpm and 80-100% tps.

And it does feel better more low down power and a cleaner more consistent torque delivery than when the staged injection passed the injectors in based on my tps and rpm.

I have a tidier ignition table I've built I've just not been brave enough to put it on with out a knock sensor. I've attached it

With regard to seeing 195ve. My brain struggles with this, my understanding must be wrong. I understood ve to be 100 when 2000cc off air would be in the engine. So without forced induction why would it be a huge amount over 100...

Thanks for your time.

While the VE table represents efficiency -- it's all relative. If every other input is not correct, then the required VE to produce the correct O2 readings may be larger (or smaller) than you would expect. For example, if the air temp measured is not the actual temp as the air goes past the valve (say it's cooler / more dense), then your VE relative to the improper air temp will be higher. If the flow of the fuel injectors does not account for the "differential pressure" between the fuel rail and the manifold air pressure, then the VE could be higher. So don't worry so much about the actual values, just give the engine what it wants to produce the desired AFR reading from the wideband lambda sensor.

My race car uses the Mazda MZR 2.0 (aka. Ford Duratec 2.0L). With AT Power throttles and a 4into1 header, my peak VE numbers were in the 115-117 range. But when I had a proper 4->2->1 exhaust system designed and built, the VE increased to about 134, and I probably added 10% power. The engine might be that much more efficient. With your highly develop engine, pistons, compression ratio, etc -- the VE number is what it is.

Your ignition table looks close to something I would run -- if you are worried about knock, take the 2000 - 2500 columns, below 50% load down a few degrees -- I find I have to use less timing there on my MZR with stock cams to avoid knock. I have a little less overall timing advance, but also am running on california premium pump fuel (91 RON=MON/2).

Okay thanks thats clearer, I have no MAP Sensor on my car. Only TPS, Water, Crank and Air Temps